Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com> writes:
> +1. It looks like a thinko from c532d15d. There's no code in between,
> so switching to oldcontext doesn't make sense.
Agreed.
> I think we also need to remove MemoryContextSwitchTo(oldcontext); at
> the end of BeginCopyTo in copyto.c, because we are not changing memory
> contexts in between.
Hmm, I think it'd be a better idea to remove the one in the middle of
BeginCopyTo. The code after that is still doing setup of the cstate,
so the early switch back looks to me like trouble waiting to happen.
regards, tom lane